For years, the problem with technology spend was not just how much was being spent. It was something more structural: IT and finance were looking at the same expenditure, but not speaking the same language.
Engineering saw services, clusters, accounts, pods, databases, and workloads. Finance saw cost centers, contracts, commitments, discounts, budgets, and period closes. Both were looking at the same bill. But not the same data structure.
That misalignment was tolerable while cloud costs could be treated as an extension of infrastructure. In 2026, it no longer is. And that is where FOCUS 1.3 changes the conversation — not because it adds "more columns," but because it makes it harder to sustain spend management based on manual translations and divergent definitions across teams.
FOCUS is no longer just a format. It is becoming the common layer for technology value
FOCUS — FinOps Open Cost and Usage Specification — defines an open, vendor-neutral schema for billing data, aligned with the FinOps Framework. Its goal: make cost information interoperable across providers and reduce operational complexity. The specification is designed to enable capabilities such as chargeback, cost allocation, budgeting, and forecasting using shared constructs regardless of where the dataset originates.
That point matters far more than it might seem. When a company works with multiple clouds, SaaS tools, data platforms, and internal systems, the challenge is no longer just visualizing spend. The challenge is normalizing it with enough consistency so that the same dataset can serve engineering, FinOps, procurement, and finance simultaneously.
FOCUS 1.3 deepens exactly that direction. The FinOps Foundation summarizes three core improvements in this version: contract commitments in a dedicated dataset, split shared cost allocation with methodological transparency, and recency and completeness signals to confirm whether data is fresh and complete before acting on it.
That is not a technical detail. It is a governance change.
What changes with FOCUS 1.3: seeing the number is no longer enough — how it was built now matters
Version 1.3 was ratified on December 5, 2025. Among its most relevant changes, it introduces a supplementary Contract Commitment dataset, new columns to expose how shared costs are split, and mandatory signals about dataset freshness and completeness. It also adds columns to distinguish between Service Provider and Host Provider — important when reseller relationships exist or services are built on top of other providers.
Previously, many organizations had to manually reconstruct three things:
Contracts existed, but they were decoupled from the cost dataset. FOCUS 1.3 introduces a separate dataset for contract commitments, with fields such as dates, remaining units, and descriptions, so that procurement, finance, and FinOps can query active commitments without mixing that information into every usage row.
In Kubernetes, shared databases, or multi-tenant services, the company could see the allocated cost but not always the methodology behind it. FOCUS 1.3 adds specific columns so that data generators can expose how costs were distributed across workloads — not just the final result.
Processing incomplete or still-provisional data breaks reconciliations and automations. FOCUS 1.3 requires timestamp, recency, and completeness signals so that users know whether the dataset is final or may still change.
That is precisely what aligns IT and finance: both stop arguing over different versions of the same data.
Why this matters more in AI than in traditional cloud
AI costs rarely arrive "clean" as a single isolated, self-contained line. They frequently appear distributed across shared compute, clusters or multi-tenant platforms, data and storage costs, variable inference, GPUs under contract commitments, SaaS layers or external APIs, and services built on top of other providers.
For this reason, AI spend tends to amplify exactly the three problems that FOCUS 1.3 sets out to solve:
- Many AI workloads run on shared infrastructure: clusters, vector databases, data services, internal platforms
- Without methodological transparency, the business sees only an inflated or arbitrary bill
- FOCUS 1.3 improves transparency around split cost allocation for provider-defined services, including containers and databases
- Reservations, consumption commitments, negotiated discounts, capacity contracts, and provider-specific agreements
- The new Contract Commitment dataset allows these terms to be separated out and queried with greater clarity
- In AI, acting on partial data is especially risky: you may optimize the wrong workload or misattribute inference costs
- FOCUS 1.3 introduces explicit signals so that users know whether the dataset is complete and up to date before processing it
As AI becomes a genuine category of technology spend, FOCUS 1.3 becomes more important, not less — because it provides a common structure precisely for the types of complexity that AI multiplies.
The real shift: FOCUS forces IT and finance to read from the same page
For years, IT could optimize infrastructure without finance truly understanding what was being optimized. And finance could demand cuts without understanding how technical costs were constructed. FOCUS 1.3 pushes organizations to resolve that divide because the conversation can now be structured around a common layer: the same column names, the same billing semantics, the same understanding of allocation, the same completeness indicators, and the same reading of contract commitments.
Saying "costs are properly allocated" is no longer enough. IT must now be able to show which method was used and on what basis costs were split across workloads.
Looking at the total is no longer enough. Finance can now understand what portion is under contract, what portion is incomplete, and what portion depends on a particular allocation methodology.
Time spent on normalization decreases and time available for analysis, prioritization, and decisions increases. This is precisely one of the FinOps Foundation's central arguments for the value of FOCUS.
The contract commitments dataset is especially useful when procurement can read it alongside actual consumption — not as separate worlds requiring manual reconciliation.
Where FOCUS 1.3 hits hardest: Kubernetes, internal platforms, and AI platforms
The split allocation changes are not minor. They hit hardest where spend is most difficult to explain: Kubernetes, shared databases, multi-tenant internal platforms, data services, and shared clusters for analytics or AI.
The FinOps Foundation itself highlights that one of the pain points FOCUS 1.3 addresses is precisely the allocation of provider-defined services that depend on containers and databases — which particularly benefits platform teams, DevOps engineers, and FinOps analysts managing multi-tenant infrastructure.
And applied to AI, the impact is even clearer:
- A single platform may serve multiple teams with undifferentiated costs
- GPU costs do not always arrive with a clear owner from the billing layer
- Inference, embedding, vector storage, and orchestration may be distributed across multiple use cases
- Without a common semantic layer, the conversation quickly degrades into political attribution arguments
FOCUS does not resolve all internal cost allocation logic on its own. But it does provide a far more robust common foundation — one that stops every conversation from having to start from scratch.
What a serious organization should do in 2026
The value of FOCUS 1.3 cannot be captured simply by "downloading a dataset." To realize its benefits, an organization should do at least four things:
The ecosystem of data generators already includes major cloud providers and several technology vendors, with support across different versions. The official site shows FOCUS dataset availability by provider and reflects a growing movement toward multi-vendor adoption.
FOCUS can serve as the common foundation, but the organization must decide how to classify and govern emerging spend categories — especially AI, where costs rarely arrive with an obvious owner.
The new transparency in FOCUS 1.3 helps, but it does not replace an internal policy for how those allocations are interpreted and validated. The standard exposes the methodology; the organization must decide whether to accept it or renegotiate it.
The contract commitments dataset is especially valuable when procurement and finance can read it alongside consumption data. Keeping it locked inside the technical stack reproduces exactly the divide that FOCUS is meant to resolve.
Yaripo's position
Most companies still treat FinOps as a cloud visibility problem. That framing is already too narrow.
In 2026, FinOps is a translation discipline — bridging technical consumption, contract logic, value allocation, budgeting, and increasingly, AI costs. FOCUS 1.3 is relevant precisely because it pushes that translation toward a common structure. It does not replace judgment. It does not replace policy. It does not replace governance. But it does reduce the baseline friction that for years forced IT and finance to work from different pages.
When AI spend begins to grow inside shared infrastructure, specific contracts, and datasets still maturing in quality, that alignment is no longer a "nice to have." It becomes a minimum condition for making sound decisions.
FOCUS 1.3 is not important because it adds more technical detail. It is important because it compels an organization to mature its understanding of technology cost. In the past, having reports was enough. Now something more is required: a shared semantic layer so that IT, finance, procurement, and FinOps stop arguing over different versions of the same reality.
And that becomes especially critical in AI, where costs rarely arrive organized, clean, or with an obvious owner. The standard does not eliminate complexity. But it finally puts it on a single, shared page.